List your company

Issue Recognition on FaultyScience (test)

95PerformanceQwen2.5-72B
6.80829.70452.675.496Mar 23, 2026

Evaluation results

MethodLinks
Qwen2.5-72BPrompt style=discriminative (P^Dis)2026.0395
Llama3.3 70BPrompt style=discriminative (P^Dis)2026.0390.8
GPT4Prompt style=discriminative (P^Dis)2026.0388.1
Mixtral-8x7BPrompt style=discriminative (P^Dis)2026.0375.2
GPT4Prompt style=generative with hint (P^Gen_+hint)2026.0348.6
Qwen2.5-72BPrompt style=generative with hint (P^Gen_+hint)2026.0340.1
GPT4Prompt style=generative without hint (P^Gen_-hint)2026.0334.9
Mixtral-8x7BPrompt style=generative with hint (P^Gen_+hint)2026.0327.8
Llama3.3 70BPrompt style=generative with hint (P^Gen_+hint)2026.0324.5
Llama3.3 70BPrompt style=generative without hint (P^Gen_-hint)2026.0314
Qwen2.5-72BPrompt style=generative without hint (P^Gen_-hint)2026.0310.8
Mixtral-8x7BPrompt style=generative without hint (P^Gen_-hint)2026.0310.2